用于人脸识别的伽柏小波联想记忆

Gabor wavelet associative memory for face recognition.

作者信息

Zhang Haihong, Zhang Bailing, Huang Weimin, Tian Qi

机构信息

Institute for Infocomm Research, Singapore 119613.

出版信息

IEEE Trans Neural Netw. 2005 Jan;16(1):275-8. doi: 10.1109/TNN.2004.841811.

DOI:10.1109/TNN.2004.841811

PMID:15732406

Abstract

This letter describes a high-performance face recognition system by combining two recently proposed neural network models, namely Gabor wavelet network (GWN) and kernel associative memory (KAM), into a unified structure called Gabor wavelet associative memory (GWAM). GWAM has superior representation capability inherited from GWN and consequently demonstrates a much better recognition performance than KAM. Extensive experiments have been conducted to evaluate a GWAM-based recognition scheme using three popular face databases, i.e., FERET database, Olivetti-Oracle Research Lab (ORL) database and AR face database. The experimental results consistently show our scheme's superiority and demonstrate its very high-performance comparing favorably to some recent face recognition methods, achieving 99.3% and 100% accuracy, respectively, on the former two databases, exhibiting very robust performance on the last database against varying illumination conditions.

摘要

这封信描述了一种高性能人脸识别系统，该系统通过将最近提出的两种神经网络模型，即加博尔小波网络（GWN）和核关联记忆（KAM），组合成一种称为加博尔小波关联记忆（GWAM）的统一结构。GWAM继承了GWN的卓越表示能力，因此展示出比KAM更好的识别性能。已进行了广泛实验，以使用三个流行的人脸数据库，即FERET数据库、奥利维蒂-甲骨文研究实验室（ORL）数据库和AR人脸数据库，来评估基于GWAM的识别方案。实验结果一致表明我们方案的优越性，并证明其高性能，与一些最近的人脸识别方法相比表现良好，在前两个数据库上分别达到了99.3%和100%的准确率，在最后一个数据库上针对不同光照条件表现出非常稳健的性能。